AI hockey predictions
AI hockey predictions24 September 2026
25 games listed; 21 include a complete forecast. See the projected winner, score, overtime probability and goal total from our ratings.
NHL NHL Preseason
FinalDallas Stars18–102Minnesota Wild16–160DAL 57%43%Projected3–2CorrectOT 19%Goals 4.6FinalAnaheim Ducks16–171Los Angeles Kings12–162ANA 59%41%Projected3–2MissedOT 17%Goals 5.423:00Carolina Hurricanes30–10Nashville Predators13–9CAR 57%43%Projected3–2OT 17%Goals 5.323:00Columbus Blue Jackets11–12Pittsburgh Penguins12–17CBJ 53%47%Projected3–2OT 18%Goals 5.123:00Detroit Red Wings8–15Buffalo Sabres22–13BUF 54%46%Projected3–4OT 16%Goals 6.123:00New York Rangers9–14New Jersey Devils12–9NYR 53%47%Projected3–2OT 18%Goals 5.023:00Philadelphia Flyers18–13Boston Bruins13–16BOS 52%48%Projected2–3OT 19%Goals 4.823:00Tampa Bay Lightning13–16Florida Panthers12–11FLA 51%49%Projected3–4OT 18%Goals 5.2
Europe DEL · Germany
Europe Elitehockeyligaen · Norway
16:30Frisk Asker9–8Narvik Hockey Arctic Eagles1–8FRA 53%47%Projected4–3OT 17%Goals 5.516:30Lillehammer IK1–7Nidaros Hockey4–4NID 56%44%Projected3–4OT 17%Goals 5.716:30Stjernen Hockey4–5Stavanger Oilers7–6STJ 52%48%Projected4–3OT 17%Goals 5.716:30Storhamar Ishockey13–4Ringerike IHK8–0RIN 50%50%Projected3–4OT 18%Goals 5.4Postp.Sparta Warriors1–5Vålerenga IF9–3VAA 57%43%Projected2–3OT 18%Goals 5.0
Europe Extraliga · Czech Republic
Europe Liiga · Finland
Europe SHL · Slovakia
Europe SHL · Sweden
17:00Djurgårdens IF4–6Växjö Lakers HC6–8DIF 61%39%Projected4–3OT 16%Goals 6.117:00Färjestad BK6–5Rögle BK12–9FBK 62%38%Projected4–3OT 16%Goals 6.117:00HV716–4Skellefteå AIK15–4SAIK 50%50%Projected3–4OT 16%Goals 6.317:00Luleå Hockey8–10Frölunda HC4–7LHF 58%42%Projected3–2OT 17%Goals 5.417:00Örebro HK7–3IF Björklöven15–2IFB 50%50%Projected3–4OT 17%Goals 6.0
Other MHL · Russia
Other NCAA Women
21:00Merrimack WarriorsLindenwood Lady Lions2–2Not enough data22:00Penn State Nittany Lions2–0Ohio State Buckeyes4–1Not enough data
How the predictions work
Each team receives a strength rating that changes after every game, based on result surprise and goal margin. A separate model measures goals scored and allowed against league averages. Home ice is estimated separately for each league. Because goals are infrequent, Poisson goal distributions generate overtime, two-goal win and total-goal probabilities. Limited game history moves teams toward the average, keeping forecasts conservative.
Each team receives a strength rating that changes after every game, based on result surprise and goal margin. A separate model measures goals scored and allowed against league averages. Home ice is estimated separately for each league. Because goals are infrequent, Poisson goal distributions generate overtime, two-goal win and total-goal probabilities. Limited game history moves teams toward the average, keeping forecasts conservative.